8D Problem Solving Process - Operational Excellence - Flevy

OPERATIONAL EXCELLENCE

C O N S U L T I N G

1 11/3/2011 – v5.0

Operational Excellence – 8D Problem Solving Process

Workshop Instructor: Frank Adler, Ph.D.

Managing Partner & LSS MBB

Operational Excellence Consulting


C O N S U L T I N G

4 11/3/2011 – v5.0

Problem Solving Process - The Terminology

� Symptom – A quantifiable event or effect, experienced by a Customer

(internal & external), that may indicate the existence of one or more problems.

� Problem – A deviation from an expectation, standard, or status quo; a

perceived gap between the existing state and a desired state. In the context of

the 8D Problem Solving Process the cause(s) of the problem is not known.

� Possible Cause – Any cause, identified through Process Mapping and C & E

Matrix, Why-Why Diagram or Fishbone Diagram, that describes how an effect

may occur.

� Most Likely Cause – A theory, based on available or collected data, that

best explains the problem description.

� Root Cause – A verified cause that convincingly supports and explains

ALL facts available and thus accounts for the problem; verified passively and

actively, by making the problem come and go. Root causes are the

fundamental, underlying reasons for a problem, e.g. management policies,

product design, system design, process capabilities, technology constraints,

standard operating procedures and human errors.

This document is a partial preview. Full document download can be found on Flevy: http://flevy.com/browse/document/opex-8d-problem-solving-process-training-module-568

http://flevy.com/browse/document/opex-8d-problem-solving-process-training-module-568


C O N S U L T I N G

7 11/3/2011 – v5.0

Problem Solving Process - The Tool Box

“If the only tool you have is a hammer, every problem looks like a nail.”

Abraham Maslow

→ You need the right tools for the right tasks.




C O N S U L T I N G

10 11/3/2011 – v5.0

The 8D (Disciplines) Problem Solving Process

1• Establish the Team

2• Describe the Problem

3• Develop Interim Containment Actions

4• Define and Verify Root Causes and Escape Points

5• Choose and Verify Permanent Corrective Actions

6• Implement and Validate Permanent Corrective Actions

7• Prevent Recurrence

8• Recognize Team and Individual Contributions




C O N S U L T I N G

13 11/3/2011 – v5.0

Problem Solving & Improvement Project Selection

� Is the problem or improvement opportunity related to the organization’sstrategic and operational objectives?

� Does the problem impact customer satisfaction?

� Does a Champion or Owner exist that is responsible for the problem?

� Do we have the resources (people, time, money) to successfully completethe project?

� Is there an objective measure or metrics in place to be improved?

� Can the project been completed in a reasonable amount of time?

� What is the likelihood of success?

� …

Every organization should develop its own Project Selection Criteria to ensure it works on the right projects at the right time.




C O N S U L T I N G

16 11/3/2011 – v5.0

Know your Strengths, but also your Weaknesses.

Solution

Task: Connect the 9 dots in the sheet of paper with exactly 4 straight lineswithout lifting your pen from the paper.




C O N S U L T I N G

19 11/3/2011 – v5.0

Basic Team Guidelines

Develop Basic Team Guidelines that the team commits to …

� Talk from personal experience and examples

� Speak to, not about, people present

� Say “I” – not “we”, “you”, “one”, “the team”, …

� Say “I would like …” or “it would help me if …” – not “you should …” or“we should …”

� Make statements before questions

� Trace opinions to observations

� Describe not judge

� Restate and build on proposals

� Recognize that feelings affect team meetings

� …




C O N S U L T I N G

22 11/3/2011 – v5.0

Purpose: Describe the internal or external problem by identifying “what is wrongwith what” and detailing the problem in quantifiable terms.

� Develop a Problem Statement

� Develop a Problem Description using the “IS – IS NOT Matrix”

� Develop a flowchart of the process and identify critical process steps with respect to the Problem Description

� Develop a Cause & Effect Matrix, Why-Why Diagram or Fishbone Diagram to identify possible causes?

� Determine whether this problem describes a “something changed” or a “never been there” situation

� Establish a high-level project plan, including milestones, project goals and objectives

“Describe the Problem” Key Activities




C O N S U L T I N G

25 11/3/2011 – v5.0

Pareto Chart & Example

� Helps to define which problems should be solved and in which order

� Presents the 20/80 – Rule

� 20% of the problems often create 80% of all the costs and/or complaints

� Analyzes defect categories over a specific time frame (e.g. 3 months)

Missing PartPosition

IncorrectPart Broken Wrong Color Stains Others

Number of Defects 40 15 10 8 6 11

Cum. Percentage 44% 61% 72% 81% 88% 100%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0

5

10

15

20

25

30

35

40

45

50

Pareto Chart - Outgoing Inspection Defects

Number of Defects

Cum. Percentage

You may need to define meaningful defect categories and categorize yourdata, before you can determine which problems to solve first




C O N S U L T I N G

28 11/3/2011 – v5.0

How to create Histogram?

Step 2: Determine the number of bars or bins to be used to create the histogramof the data points. Calculate the width of one bar by dividing the range of yourdata by the number of bars selected.

Number of Bars:

less than 50

50 - 100

100 - 250

over 250

5 or 7

5, 7, 9 or 11

7 - 15

11 - 19

Number of Data Points:

Minimum = 2.1

Maximum = 3.1

Range = 1.0

Bar Width = 0.2 (5 Bars)




C O N S U L T I N G

31 11/3/2011 – v5.0

Histogram – Different Patterns of Variation

1. The bell-shaped distribution:

Symmetrical shape with a peak in the middle of therange of the data.

While deviation from a bell shape should beinvestigated, such deviation is not necessarily bad.

2. The double-peaked distribution:

A distinct valley in the middle of the range of the datawith peaks on either side.

This pattern is usually a combination of two bell-shapeddistributions and suggests that two distinct processes(e.g. machine, operator, method, material, …) are atwork.




C O N S U L T I N G

34 11/3/2011 – v5.0

The 5 Why’s

Description: The 5 Why’s Analysis helps to identify the final problem statement or in

some cases even the root cause. It helps to distinguish between the symptom of a

problem and the problem itself and it encourages the team to reach an answer that is

fundamental and actionable.

Procedure:

Step 1: Develop an initial problem statement of the specific problem to be solved. Write it

in the upper left corner of a piece of paper, flip chart or white board.

Step 2: Ask “Why?” this problem does or could occur. Write the cause (object & defect)

underneath the initial problem statement.

Step 3: The cause identified in Step 2 now becomes a new problem statement. Repeat

Step 2 and ask “Why?”, e.g. “Why would this situation occur?", again.

Step 4: Continue Step 2 and Step 3 until you cannot for certain answer the question or

reach an answer that is fundamental and actionable.This document is a partial preview. Full document download can be found on Flevy: http://flevy.com/browse/document/opex-8d-problem-solving-process-training-module-568



C O N S U L T I N G

37 11/3/2011 – v5.0

Problem Statement – Example 1

What is the Problem Statement (Object & Defect)?

WHY? → The tool was not in inventory.

WHY? → The Supplier did not deliver the replacement tools.

WHY? → The Supplier did not receive our Purchase Order on-time.

WHY? → We did not submit the Purchase Order on time.

WHY? → We did not have the “right” rework tool.

WHY? → We don’t know.

Initial Problem Statement

Final Problem Statement




C O N S U L T I N G

40 11/3/2011 – v5.0

Case Study Group Work 1

The Task:

� Read the introduction to the “Felt Tip Marker” Case Study

� Develop the Problem Statement for the Case Study

5 Minutes




C O N S U L T I N G

43 11/3/2011 – v5.0

Swimlane Process Flowchart

…

A swimlane is a visual element used in process flowcharts that visuallydistinguishes responsibilities for activities or sub-processes of a businessprocess. Swim lanes may be arranged either horizontally or vertically.




C O N S U L T I N G

46 11/3/2011 – v5.0

The Problem Description

The Problem Description defines the boundaries of the problem, in terms of what it is

and what it is not but could be.

Scientific description of any event can be made by providing information on

� What the problem is and what it is not but could be?

� IS - The long brackets are missing,

� IS NOT - The short brackets are not missing (but could be)

� Where the problem is and where it is not but could be?

� IS - The order entry defects are in direct ship orders

� IS NOT - The order entry defects are not in stock orders (but could be)

� When the problem occurs and when it does not but could?

� IS - The clips started breaking in early 2010

� IS NOT – The clips did not break before early 2010 (but could have)

� How big the problem is and how big it is not but could be?

� IS - About 50% of the paychecks

� IS NOT – Not all paychecks are incorrect (but could be)




C O N S U L T I N G

49 11/3/2011 – v5.0

The “IS – IS NOT” Matrix

The “IS – IS NOT” Matrix in the OpEx Problem Solving Process Worksheet

identifies where to start looking for possible causes by isolating the what, when, where

and how big about the problem statement.

Step 1: Enter the Problem Statement.

Step 2: Use the “IS” column to describe what objects are effected and what the defect

is. Use the “IS NOT” column to describe what could be effected and occur but does not.

Step 3: Use the “IS” column to describe where the event occurs. This can be

geographical, physical, or on an object. Use the “IS NOT” column to describe where the

problem could occur but does not.

Step 4: Use the “IS” column to describe when the problem occurs. When did it happen

first? What pattern of reoccurrence? Also, when the event occurs in relation to other

events. Use the “IS NOT” column to describe when the problem could occur but does

not.

Step 5: Use the “IS” column to describe the extent of the problem. How many objects

or occurrences had problems? How many problems? How serious are they? Use the

“IS NOT” column to describe how big or small the problem could be but is not.




C O N S U L T I N G

52 11/3/2011 – v5.0

Cause & Effect Matrix – Phase 1

2. Score (0 – 100) the Importance or Criticality of each Problem Statement, with the total score equal to 100

3. List the Process Steps or Activities

4. List Inputs for each Process Step

1. List the Problem Statement(s)

We will cover this later.This document is a partial preview. Full document download can be found on Flevy: http://flevy.com/browse/document/opex-8d-problem-solving-process-training-module-568



C O N S U L T I N G

55 11/3/2011 – v5.0

Brainstorming is used to generate a large number of ideas in a short period of

time.

� Step 1: Review the topic or problem to be discussed. Make sure the entire

team understands the subject of the brainstorm. Clarify if you are looking

for possible causes or possible solutions.

� Step 2: Allow a couple of minutes for everyone to think about the task and

write down some ideas.

� Step 3: Invite the team members to call out their ideas, randomly or in

turns around the table. Make sure that the team builds upon each others

ideas (1 + 1 > 2), allow clarifications but avoid discussions or evaluations.

� Step 4: Record all ideas, e.g. on post-its, and stick them on a

flip chart visible to everyone.

� Step 5: Continue with Step 3 and 4 until several minutes

silence produces no more ideas.

Brainstorming – Process




C O N S U L T I N G

58 11/3/2011 – v5.0

Fishbone Diagram – Phase 1

Alternative 2 (Team-defined Categories)

Step 1: Draw a horizontal trunk line and to

the right end of this line write the

Problem Statement.

Step 2: Brainstorm for all possible causes.

Step 3: Group the possible causes into

categories and give each category a

name (min. 2, max. 6 categories).

Step 4: Draw a main branch for each

category and put a category name at

the end of each branch.

Step 5: Add minor branches and the

identified possible causes under

each major branch.

Step 6: Check that the diagram is complete

and logical.

Alternative 1 (Pre-defined Categories)

Step 1: Draw a horizontal trunk line and

to the right end of this write the

Problem Statement.

Step 2: Draw the main branches and put

the name of the main categories,

e.g. 4Ms & 1E, at the end of the

main branches.

Step 3: Brainstorm for all possible causes

Step 4: Add minor branches and the

identified possible causes under

each major branch.

Step 5: Check that the diagram is

complete and logical.




C O N S U L T I N G

61 11/3/2011 – v5.0

“Phase D2” Checklist Questionnaire

Problem Description (cont.):

� Determination has been made as to whether this problem describes a“Change-Induced” or a “Day-One-Deviation” situation?

� The current process flow has been identified and documented (asneeded)?

� Process detail has been reviewed to determine where this problem first appears?

� Problem pattern(s) has been considered / evaluated?

� Similar components and/or parts have been reviewed for the same problem?

� All required data has been collected and analyzed?

� A Cause & Effect Matrix, Why-Why Diagram or Fishbone Diagram has been developed to identify possible causes?




C O N S U L T I N G

64 11/3/2011 – v5.0

“Develop Interim Containment Actions” Key Activities

Purpose: Define, verify and implement Interim Containment Action toisolate the effects of the problem from any internal and/or externalCustomer until Permanent Corrective (Preventive) Action are implemented.

� Define potential Interim Containment Action

� Verify effectiveness of potential Interim Containment Action

� Select and implement Interim Containment Action

� Validate effectiveness of implemented Interim Containment Action with the Customer




C O N S U L T I N G

67 11/3/2011 – v5.0

PDCA or Deming Management Cycle

The “Develop Interim Containment Actions” Phase follows the PDCA Cycle

1. Plan

(Plan & Re-Plan)

2. Do

(Implement)

3. Check

(Monitor)

4. Act

(Evaluate)

Deming

PDCA Cycle

Determine which steps must be completed to achieve the plan’s objectives by the target date.

Implement and manage the implementation plan.

Record the results of the implementation phase.

Evaluate the results and decide if additional actions need to be developed.




C O N S U L T I N G

70 11/3/2011 – v5.0









The 8D Problem Solving Process




C O N S U L T I N G

73 11/3/2011 – v5.0

Comparative Analysis - Example

� Problem Description:

� IS – The Red Makers leak

� IS NOT – The Blue Markers could leak, but are not

� Comparative Analysis:

� Distinctions

� 1. Different Color Plastic,

� 2. Different Color Ink,

� 3. Strawberry Scent in Red Marker

� Changes

� 1. New Plastic Supplier (5 weeks ago),

� 2. Added Powder to Red Marker (2.5 weeks ago)




C O N S U L T I N G

76 11/3/2011 – v5.0

Cause & Effect Matrix – Phase 2

5. Score Correlation

between Process

Inputs and Problem

Statement or Defect3. List the Process Steps or Activities

4. List Inputs for each Process Step

1. List the Problem Statement(s)

2. Score (0 – 100) the Importance or Criticality of each Problem Statement, with the total equal to 100




C O N S U L T I N G

79 11/3/2011 – v5.0

Fishbone Diagram – Phase 2

In Phase 2 of the Fishbone Diagram, the team tries to reduce the large number ofpossible causes to a short list of most likely causes, using the Problem Description,data analysis and the team’s combined process knowledge and experience.

Remember, the root cause(s) convincingly supports and explains ALL facts availableand thus accounts for the problem. So, you need to ask yourself “If this would bethe root cause of the problem, would it explain the Problem Description?”

4M’s = MEN

METHODS

MATERIALS

MACHINES

+ ENVIRONMENT

4P’s = PEOPLE

PLANT

POLICIES

PROCEDURES

Machines

Materials Methods

Environment

Trunk

Primary Possible Cause Category

Main Branch

Minor Branch

Second-Level

Possible Causes

Men/People

Problem Statement

Most Likely

Cause #1

Most Likely

Cause #3

Most Likely

Cause #2




C O N S U L T I N G

82 11/3/2011 – v5.0

Scatter Plot – A little Bit of Statistics

� With a little bit of statistics we can build a simple equation or model topredict the “Ice Cream Sales (in $)” for every “Temperature (in ºC)”.

Ice Cream Sales (in $)” = m � Temperature (in ºC) + c

� To determine the parameters “m” and “c”, we need two data points (x → y)from our scatter plot, for example (13 ºC → $205) and (22 ºC → $500).

� � ��

� �� How much “Ice Cream Sales (in $)” is our model predicting when the

“Temperature (in ºC)” is 19 ºC ?




C O N S U L T I N G

85 11/3/2011 – v5.0

Root Cause Theory Testing

PROBLEM SOLVING PROCESS WORKSHEET

Problem Statement

(What is wrong with what?):Red Felt Tip Marker Leaks

IS IS NOT DISTINCTIONS CHANGESTEST FOR MOST LIKELY CAUSES

(+) (-) (?)

Describe what does occurDescribe what does not occur, but

could occur

What could explain the IS-IS NOT?

(People, Methods, Material,

Machines, Environment)

What has changed in, on, around or

about this distinction? When

did it change?

<Define Possible Root Cause

Theory>

What?

Object:

Defect:

Where?

Where is the object when the

defect is observed? Where is the

defect on the object? Are there

any pattern or trends identifiable?

When?

When was the defect observed

first? Can or was the defect be

observed before, during or after

other events? When since has the

defect occurred? Are there any

pattern or trends identifiable?

How Big?

How many objects have the

defect? What is the size of a single

defect? How many defects are on

each object? Are there any

pattern or trends identifiable?

2

3

3

3

3+ - ?




C O N S U L T I N G

88 11/3/2011 – v5.0

Root Cause Verification

� Active Verification is a process where …

� … the problem solver uses the variable thought to be the root cause tomake the effect come and go. Both coming and going are important teststo confirm root causes.

� Passive Verification is done by observation.

� With passive verification, you look for the presence of the root causewithout changing anything. If you cannot prove the presence of the rootcause, then chances are great that this most likely cause is not the rootcause.

Step 1: Identify, document and perform a passive and/or active

root cause verification method.

Step 2: Continue with the 8D Problem Solving Process if the

passive and/or active root cause verification was successful.

Otherwise, develop and test additional Root Cause Theories.




C O N S U L T I N G

91 11/3/2011 – v5.0


Escape Points:

� The process flow was reviewed and control points associated with theroot cause(s) identified.

� A determination was made as to the existence of a control system todetect the problem?

� If a control system exists, changes from its original design (if any) havebeen identified?

Project Management:

� The Project Champion has reviewed and supports the analysis, findings and conclusions?

� 8D Problem Solving Process Worksheet & Project Documentation has been updated?




C O N S U L T I N G

94 11/3/2011 – v5.0

Many Human Errors are designed into the Process

Reducing & Eliminating

Human Error needs more

than additional Training




C O N S U L T I N G

97 11/3/2011 – v5.0

6. Incorrect Selection (or Identification)

→ What object of the process is prone to be incorrectly selected or identified?

7. Incorrect Counting (or Calculating)

→ What objects of the process can be counted, measured or calculated incorrectly?

8. Misrecognition (or Misunderstanding or Misreading)

→ What misunderstanding or misreading is prone to occur?

→ What information, risk or failure/error is prone to be overlooked?

→ What miscommunication is prone to occur?

→ What incorrect decision is prone to occur?

Sixteen Human Error Modes




C O N S U L T I N G

100 11/3/2011 – v5.0

Six Mistake Proofing Principles

1. Elimination seeks to eliminate an error-prone process step by redesigning the product

or process so that the task or part is no longer necessary.

Example: An example of elimination is the use of ambient-light sensors to turn outside

lighting on and off.

2. Prevention modifies the product or process so that it is impossible to make a mistake

or that a mistake becomes a defect.

Example: An example would be the implementation of an automatic dispenser to insure

the correct amount of adhesive is applied during an assembly process or the coin

dispenser in food stores preventing that customers are getting short changed.

3. Replacement substitutes a more reliable process to improve repeatability. This

includes use of robotics or automation that prevents a manual assembly error.

Example: An example would be a camera that will not function when there is not enough

light to take a picture. Also some clothes dryers shut down when they detect an

overheating situation.




C O N S U L T I N G

103 11/3/2011 – v5.0

Poka (= inadvertent error) - Yoke (= avoid)

devices help us avoid defects, even when inadvertent errors are made.

Poka - Yoke Devices help build Quality

into Processes and Products

Human Errors - Poka-Yoke Mistake Proofing

The term Poka-yoke (poh-kah yoh-keh) was coined inJapan during the 1960s by Shigeo Shingo who was oneof the industrial engineers at Toyota.

Toyota: 80-90% of problems

are resolved on the floor




C O N S U L T I N G

106 11/3/2011 – v5.0

Seven Best Poka-Yoke Devices – Guide Pins

1. Guide Pins

Guide pins of different sizes and/or shapes and placed in the proper locations ensure

that parts are being assembled correctly by providing the operator feedback when a

mistake has been made. Guide pins can also be used in jigs to ensure proper

positioning of the part.

Applications

• Proper alignment of a work piece

• Proper orientation of a work piece

Features

• Easy to develop & implement

• May be the result of DFA and DFM activities

(Product Quality Planning)

Human Error Prevention

• wrong order, incorrect selection, incorrect positioning, incorrect orientation, …




C O N S U L T I N G

109 11/3/2011 – v5.0

4.1 Proximity Sensors

Proximity sensors emit a high-frequency magnetic field and detect an upset in the field

when an object enters it. They can be used to detect the presence or absence of an

object.

Applications

• Sensing of tank or bin level

• Confirmation of part or object passes by

• Detection presence or absence of object

• Positioning of work piece

Features

• Non-contact - Work in harsh environments - Small sensors are available for

installation in tight areas - Fast response speed


• Omission, excessive/insufficient repetition, incorrect selection, incorrect counting,

incorrect positioning, incorrect orientation, …

Seven Best Poka-Yoke Devices – Sensors




C O N S U L T I N G

112 11/3/2011 – v5.0

6. Counters & Timers

Counters (optical or electro-mechanical) look at the occurrence of events. They are usually

triggered by some type of sensor. Counters can be programmed to shut down the process

if a set number of events do not occur or if too many events do occur. Timers can shut

down the process if processing time or activity time does not meet or exceeds a preset

level.

Applications

• Ensuring the proper number of events occurred

• Preventing failure of equipment or a component

by timing usage

Features

• Flexible - Easy to use - Easy for people to understand


• excessive/insufficient repetition, incorrect counting, incorrect positioning, incorrect

orientation, …

Seven Best Poka-Yoke Devices – Counters & Timers

Correct Number of HolesThis document is a partial preview. Full document download can be found on Flevy: http://flevy.com/browse/document/opex-8d-problem-solving-process-training-module-568



C O N S U L T I N G

115 11/3/2011 – v5.0

Human Errors → Mistake-Proofing Examples

Never forget your cell phone again.

«Pick to Light»: on an assembly line, if the above lightis green means that you must take the piece. If thelight above the rack is red means that you must nottake that piece.




C O N S U L T I N G

118 11/3/2011 – v5.0


An oil change facility puts the dip stick on thefender protector.

Removing the fender protector will cause thedip stick to clatter on the floor unless it has beenreinserted. → Workplace Organization

After a patient died from receiving a medicationthat was not properly diluted, all of that medicationwas diluted before being stored. → Pre-Kitting




C O N S U L T I N G

121 11/3/2011 – v5.0


Consumer friendly Mistake-Proofing productdesign improves usability and CustomerSatisfaction.

… and many more.

Preventing missing weld nuts, with a sensorlinked to a visual & audio alarm. Process willstop automatically and a corrective action isrequired.




C O N S U L T I N G

124 11/3/2011 – v5.0

Solution Selection Matrix – Basic Example

• Effectiveness Effectiveness of the solution to resolve the root cause or escape

point of the problem

• Cost in local currency Cost to implement the solution

• Time in weeks Time to implement the solution

• Resources Yes = enough resources (own)

Partly = not enough own resources, but more available in the

organization

No = no resources available in the organization

SOLUTION # 1

SOLUTION # 2

SOLUTION # 3

SOLUTION # 4

Effectiveness Cost Time Resources

70 %

90 %

50 %

65 %

20 000

52 000

24 000

18 000

14

5

12

15

Yes

Partly

No

Yes




C O N S U L T I N G

127 11/3/2011 – v5.0













C O N S U L T I N G

130 11/3/2011 – v5.0

Verify Solution Effectiveness

� Does the data indicate that the problem has been solved or thatperformance has improved?

� Do changes in performance align with improvement actions taken?

94.0%

94.5%

95.0%

95.5%

96.0%

96.5%

97.0%

97.5%

98.0%

98.5%

99.0%

On-Time Delivery 2012 (Direct & Non-Stock Orders)

Series1

Implementation of Solution A

Implementation of Solution B




C O N S U L T I N G

133 11/3/2011 – v5.0

B vs. C Experiment

� In the B vs. C experiment, the symbols “B” and “C” stand for twoprocesses or methods or policies that need to be compared.

� “C” is, generally, the current process or method, and “B” is, supposedly,a better process or method.

� The power of the B vs. C experiment is that often with just a few units -most frequently, three “B” units and three “C” units - we can determinewhether the two process, methods or policies are leading to differentresults or not.

� A unit can consist of only one product or a small batch of products, usingthe average of each batch as the result for further analysis.

B B B C C C




C O N S U L T I N G

136 11/3/2011 – v5.0

How to Perform a B vs. C Experiment?

Step 2: Randomize and Conduct the Experiment

Randomizing the sequence of testing, say, three “B” units and three “C” units, is amust.

Randomizing filters out other variables, so that the test results can be correctlyattributed to the potential root cause - namely, the variable altered from “C” to “B”.

It is tempting, in production, to run three “C” units with the existing process andthen switching over to the “B” process to produce the “B” units. But that woulddefeat randomization, and the results could be invalid.




C O N S U L T I N G

139 11/3/2011 – v5.0

B vs. C Experiment – Exercise

The “Company XYZ Screw Production Process”

Task: Perform a B vs. C experiment to convince the company XYZ’s management team

that the new process will indeed produce shorter screws and resolve the received

Customer complaints.

Conduct Experiment & Collect Data:

Analysis & Conclusion:




C O N S U L T I N G

142 11/3/2011 – v5.0













C O N S U L T I N G

145 11/3/2011 – v5.0

Standard Work – Why should we bother?

We need to do away with the notion that standards necessarily mean rigidity.Rather, standard work can help people do their jobs consistently andreliably, and then improve how they do it.

The traditional view that efficiency requires bureaucracy and that bureaucracyimpedes flexibility needs to be replaced with a new model: clever applicationof standard work allows you to have efficiency and flexibility.

Our ability to improve the ways we do things depends on defining andshaping our daily habits of mind and practice — our "standard work.“




C O N S U L T I N G

148 11/3/2011 – v5.0

Process Control Plan – Objective

A Control Plan is a written statement of an organization’s quality planningactions for a specific process, product, or service.

The Objective of an effective Process Control Plan is to

� operate processes consistently on target with minimum variation,which results in minimum waste and rework

� assure that product and process improvements that have beenidentified and implemented become institutionalized

� provide for adequate training in all standard operating procedures,work instructions and tools

CustomerRequirements

Product & PartCharacteristics

ProcessInput & OutputCharacteristics

ProcessControls

ProcessControl

Plan




C O N S U L T I N G

151 11/3/2011 – v5.0

� Sample Size: Sample size specifies how many parts are evaluated at anygiven time. The sample size will be “100 %” and the frequency“continuous” in case of 100% inspection.

� Sample Frequency: Sample frequency specifies the how often a samplewill be taken, e.g. once per shift or every hour.

� Control Method: Brief description of how the information/data will becollected, analysed/controlled and reported. More detailed informationmay be included in a named work instruction.

� Reaction Plan: Necessary corrective actions to avoid producing non-conforming products or operating out-of-control. Corrective actions shouldnormally be in the responsibility of the person closest to the process, e.g.the machine operator. This is to secure, that immediate corrective actionswill take place and the risk of non-conforming products will be minimized.More detailed information may be included in a named work instruction.

Process Control Plan – Template




C O N S U L T I N G

154 11/3/2011 – v5.0

� The OCAP is a systematic and ideal problem-solving tool forprocess problems because it reacts to out-of-control situations inreal time.

� OCAPs standardize the best problem-solving approaches from themost skilled and successful problem solvers (experts/operators).

� The OCAP also allows (and requires) off-line analysis of theterminators to continually improve OCAP efficiency.

Some Benefits of Out-of-Control-Action-Plans




C O N S U L T I N G

157 11/3/2011 – v5.0


Corrective & Preventive Solutions (cont.):

� Plans have been written to coordinate preventive actions andstandardize the practices?

� A method for communicating to those affected by the changes in thenew practices has been determined?

� The Champion concurs with the identified corrective & preventiveactions and plans?

Project Management:

� 8D Problem Solving Process Worksheet & Project Documentation has been updated?




C O N S U L T I N G

160 11/3/2011 – v5.0

8D Problem Solving Process Report

Microsoft Word

Document




C O N S U L T I N G

163 11/3/2011 – v5.0

Additional Material



1

Flevy (www.flevy.com) is the marketplace

for premium documents. These

documents can range from Business

Frameworks to Financial Models to

PowerPoint Templates.

Flevy was founded under the principle that

companies waste a lot of time and money

recreating the same foundational business

documents. Our vision is for Flevy to

become a comprehensive knowledge base

of business documents. All organizations,

from startups to large enterprises, can use

Flevy— whether it's to jumpstart projects, to

find reference or comparison materials, or

just to learn.

Contact Us

Please contact us with any questions you may have about our company.

• General [email protected]

• Media/[email protected]

• [email protected]

8D Problem Solving Process - Operational Excellence - Flevy

Documents